A KEYHOLE PREVIEW OF THE LAWS FOR AI

INDIA HAS TO FIND ITS FOOTHOLD IN EVOLVING LAWS THAT ALIGN WITH SECURITY AND PROTECTIVE NEEDS AND ALSO IN FACILITATING BUSINESS GROWTH BY FOSTERING CERTAINTY AND TRUST

- 

“Real or fake?” The oft-repeated question boomed across households glued to stories of obscene or defamatory celebrity deepfakes. ChatGPT had already made Artificial Intelligence dinner table conversation-worthy. Copyright and personality violation claims, misinformation, defamation, and fabricated cases being presented due to “AI hallucinations” have led to several cases being filed against generative AI developers. With cases of AI abuse capturing headlines, it is no surprise that the clamour for strict laws or regulations to curb abuse have taken centre stage. The positive use of AI runs the risk of being lost in this noise. Clearly, a legislative tightrope awaits us—one that calls for balancing innovation, growth and harnessing economic benefits with combating crimes or violations.

The European Union’s AI Act— over two years in the making and realised in December 2023—represents one of the first standalone legislations for this domain. Touted as a comprehensive legislation for AI, the Act takes a risk-based approach to ban what are listed as unacceptable risk models and regulate high-risk categories. Those perceived to be low risk are left alone for self-regulation. The US is evaluating two federal laws for combating deepfakes and misinformation campaigns with the DEEPFAKES Accountability Act and PROTECT Act. Singapore’s Protection from Online Falsehoods and Manipulation Act 2019, as the name suggests, combats misinformation and disinformation campaigns intended to harm inter alia